Resolve 17 to Youtube HDR

Get answers to your questions about color grading, editing and finishing with DaVinci Resolve.
  • Author
  • Message
Offline

thefilmaddict

  • Posts: 129
  • Joined: Sun Oct 27, 2019 9:58 pm
  • Real Name: Jason Morris

Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 1:48 pm

I just edited and colored a short video in HDR using resolve 17, studio 4k mini and an ASUS 1000 nit monitor. The video looks great while editing. When I export it, it's no longer in HDR and looks terrible (crushed blacks, too much chroma and blown out highlights). Youtube does not recognize it as HDR after I upload it (usually when it does, HDR comes up on the side of the screen). I think that I am using the wrong delivery/render settings. Does anyone know the right ones to use? There are videos out there with suggestions, but none that I've run into that mention the latest version of resolve (the new version has slightly different settings). Thanks for your help!
Offline

thefilmaddict

  • Posts: 129
  • Joined: Sun Oct 27, 2019 9:58 pm
  • Real Name: Jason Morris

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 4:14 pm

I used this video to setup my project. It features a Blackmagic employee.

Offline

thefilmaddict

  • Posts: 129
  • Joined: Sun Oct 27, 2019 9:58 pm
  • Real Name: Jason Morris

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 4:35 pm

From YouTube: If you're grading your video, grade in Rec. 2020 with PQ or HLG.
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 7:49 pm

Mentioned video may be fine for monitoring needs (actually Enable HDR metadata over HDMI is not ticked, so not even sure about correctness of this video at all), but for your YT needs settings are wrong.

A lot of presets in 17.3 is simply wrong.
For example HDR Rec.2020 doesn't limit gamut to P3, which you would really like. HDR P3 Broadcast is not very good either as it uses P3 D65 Intermediate color timeline. Better to use 2020 and then limit to P3 I assume. Very confusing presets, which hide all detailed info and mislead you.

Same with 17.4 Automatic color management.
Color Processing Mode= HDR
Output Color Space = HDR PQ
Is gamut limited to P3? (if you untick auto and then go to custom mode you will find out it's NOT)

About any delivery requires Rec.2020 (because P3 is not a 'transmission' standard), so actual used gamut should be clearly stated:
Output Color Space:
    HDR PQ (P3 limited)
    HDR PQ (Rec.2020)
    HDR HLG(P3 limited)
    HDR HLG (Rec.2020)

17.4 with unticked Automatic color management has 'best' presets.
You need to use this:
Screenshot 2021-10-27 at 21.28.04.png
Screenshot 2021-10-27 at 21.28.04.png (94.15 KiB) Viewed 3673 times

plus tick Enable HDR10+ (I have no studio version, so not showing up) and set 'HDR mastering is for' 1000 nits.
Then during export tick Embed HDR10 Metadata (although Resolve still doesn't write MaxFALL and MaxCLL if I'm right).

If you have no Studio then you can't tick these options anyway, but still YT will detect file as HDR (and use some default values for HDR metadata, which should be ok). Key trigger for YT to detect file as HDR is output colro space- it has to be Rec.2020 PQ/ST2084 (or HLG).

HDR presets are still far from optimal. BM needs to sit (with experienced colourists as well) and sort it out once for good (so less experienced users have some good templates). So far it has been chaotic jumping around all possible options.
Last edited by Andrew Kolakowski on Wed Oct 27, 2021 8:20 pm, edited 1 time in total.
Offline

Jason Conrad

  • Posts: 797
  • Joined: Wed Aug 16, 2017 3:23 pm

Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 8:14 pm

It’s been a few years since I messed with it, but one of the larger hurdles was that not all YouTube players were capable of recognizing HDR metadata flags. I think Chromecast 4K was one of the few devices containing a YouTube player that would work. TVs were hit or miss, depending on their android version.


Sent from my iPad using Tapatalk
-MacBook Pro (14,3) i7 2.9 GHz 16 GB, Intel 630, AMD 560 x1
-[DR 17.0 Beta9]
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 8:22 pm

First you make sure YT itself recognises uploaded file as HDR. Players behaviour is out of your control and it has not much to do with your uploaded files as YT re-encodes everything.
Offline

Jason Conrad

  • Posts: 797
  • Joined: Wed Aug 16, 2017 3:23 pm

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 8:25 pm

Also, I feel like Apple is largely to blame over P3/ “wide gamut” shenanigans, but BMD doesn’t help by catering to Apple’s marketing BS. Same thing goes for rec.709-A.


Sent from my iPad using Tapatalk
-MacBook Pro (14,3) i7 2.9 GHz 16 GB, Intel 630, AMD 560 x1
-[DR 17.0 Beta9]
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 8:48 pm

P3 has nothing to do with Apple. Apple adopted it as well instead of limiting (very capable) current displays to smaller Rec.709.
It's widest defined gamut current displays can show, so it has been adopted.
Rec.2020 atm. is just a theory.
P3 existed for ages and been used in cinema. Now it 'moved' to home as well.
Last edited by Andrew Kolakowski on Wed Oct 27, 2021 8:52 pm, edited 1 time in total.
Offline

Jason Conrad

  • Posts: 797
  • Joined: Wed Aug 16, 2017 3:23 pm

Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 8:51 pm

Andrew Kolakowski wrote:P3 has nothing to do with Apple.
Exactly. Now, if somebody would just tell Apple that.


Sent from my iPad using Tapatalk
-MacBook Pro (14,3) i7 2.9 GHz 16 GB, Intel 630, AMD 560 x1
-[DR 17.0 Beta9]
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 9:08 pm

Not sure what is your problem with P3 and Apple.
Fact that you should limit gamut to P3 in Resolve has nothing to do with Apple.
It's a defined gamut which for long time wasn't really achievable at home (same as Rec.2020 isn't now). It has been used in cinemas as projectors were able to cover it. It was treated as high-end option.
Once technology got there companies started adopting it in consumer products as well.
Apple uses good screens in their hardware, so they were one of the first to move to P3 screen opposite or sRGB (which is same as Rec.709). No idea why would you put good/expensive screens and then stayed at sRGB gamut- pointless.
Today good TVs can also about cover P3 gamut, so content is mastered to P3.
Problem is that P3 gamut is not defined in any spec as 'transmission' gamut.
When your TV talks to eg. BD player it establishes connection as Rec.709 for HD or Rec.2020 for UHD. There is no P3 option defined there. This is why we 'insert' P3 gamut into Rec.2020, so signal is Rec.2020, but in reality colors are limited to P3 gamut. This is also why HDR spec has mastering display metadata which tells TV about real gamut of the video, so it doesn't blindly try to always use Rec.2020.
If you were to use Rec.2020 then you would have massive variation on TVs as some support just 40% of Rec.2002, others up to 80%. This is why we need to limit gamut to P3 which is typically covered in 90+%, so at the end you don't see much of a difference. Difference on TV which supports 90 vs. 99% of P3 is not very visible (we still have plenty for monitors which cover just 60% of sRGB!). In case of eg. 40% vs. 80% for Rec.2020 you would see a drastic difference.
Rec.2020 is now what P3 was long time ago. Wide, theoretical gamut which is not achievable at home.
Latest (crazy expensive) laser projectors can fully/about hit Rec.2020, so soon we may see mastering to Rec.2020 for cinema needs. Yet again cinema will stand out from home experience.
Offline

Jason Conrad

  • Posts: 797
  • Joined: Wed Aug 16, 2017 3:23 pm

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 9:45 pm

Andrew Kolakowski wrote:Not sure what is your problem with P3 and Apple.
Fact that you should limit gamut to P3 in Resolve has nothing to do with Apple.
It's a defined gamut which for long time wasn't really achievable at home (same as Rec.2020 isn't now). It has been used in cinemas as projectors were able to cover it. It was treated as high-end option.
Once technology got there companies started adopting it in consumer products as well.
Apple uses good screens in their hardware, so they were one of the first to move to P3 screen opposite or sRGB (which is same as Rec.709). No idea why would you put good/expensive screens and then stayed at sRGB gamut- pointless.
Today good TVs can also about to cover P3 gamut, so content is mastered to P3.
Problem is that P3 gamut is not defined in any spec as 'transmission' gamut.
When your TV talks to eg. BD player it established connection as Rec.709 for HD or Rec.2020 for UHD. There is no P3 option defined there. This is why we 'insert' P3 gamut into Rec.2020, so signal is Rec.2020, but in reality colors are limited to P3 gamut.
If you were to use Rec.2020 then you would have massive variation on TVs as some would support just 50% of Rec.2002, others up to 80%. This is why we need to limit to P3 which is typically covered in 90+%, so at the end you don't see much of a difference. Difference on TV with supports 90 and 99% of P3 is not very visible. In case of eg. 50% and 80% for Rec.2020 you would see drastic difference.
Apple has oversold the capability of their monitors since at least 2017, calling them “wide gamut” and shipping them with .icc profiles named “display P3,” which doesn’t cover the P3 gamut used in theaters, nor can those monitors cover that gamut. Maybe they do now, but the one sitting in front of me doesn’t.

Same goes for brightness. Apple calls it “wide gamut” knowing that consumers will falsely assume HDR. Their marketing spin is playing out again with “XDR,” which isn’t an actual standard, so who knows what it really is? Until the new machines are in the wild and measured against actual standards, “XDR” is meaningless (even against their previous Cinema Display), and is a shameless money grab.


Sent from my iPad using Tapatalk
-MacBook Pro (14,3) i7 2.9 GHz 16 GB, Intel 630, AMD 560 x1
-[DR 17.0 Beta9]
Offline
User avatar

AlexWalker47

  • Posts: 25
  • Joined: Wed Dec 18, 2019 6:52 pm
  • Real Name: Alex Walker

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 11:00 pm

www.alexwalker.co
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 11:12 pm

Jason Conrad wrote:Apple has oversold the capability of their monitors since at least 2017, calling them “wide gamut” and shipping them with .icc profiles named “display P3,” which doesn’t cover the P3 gamut used in theaters, nor can those monitors cover that gamut. Maybe they do now, but the one sitting in front of me doesn’t.

Same goes for brightness. Apple calls it “wide gamut” knowing that consumers will falsely assume HDR. Their marketing spin is playing out again with “XDR,” which isn’t an actual standard, so who knows what it really is? Until the new machines are in the wild and measured against actual standards, “XDR” is meaningless (even against their previous Cinema Display), and is a shameless money grab.


Sent from my iPad using Tapatalk



No idea why "wide gamut" brings people to assumptions it's HDR screen. 1min of google will tell you gamut and brightness aren't always directly linked, so such an assumption is very wrong. Sorry, but blame yourself to be too lazy to spent 1 min on internet to verify Apple PR crap. Same with XDR. They just made it up to sound nice and show they are better than HDR standard. No one can blame them to give their screen/technology own name.

There is only one P3 gamut.
Maybe old screens were not covering whole P3 gamut (but as soon as you pass 90% it's not bad).
Quick internet search suggests that even 2015 screens were around 90%+ of P3, so this is fine for home device.
If you talking about articles like this:
https://blog.conradchavez.com/2015/10/2 ... late-2015/
then there are some "strange assumptions" there. All what Apple does is use P3 gamut as a reference. If someone says it's not the same as 'real' P3 DCI standard then it just shows lack of understanding the matter.
All what Apple needs from DCI P3 standard is P3 gamut. Rest has absolutely no usage for home needs (do you want to have 48nits brightness and 2.6 gamma?). All what Apple said is that they use wide gamut screens (P3 gamut with D65 white point) and the rest are just assumption made up by "many internet experts". You have many complains that those screens are not good to do real DCI work (did anyone including Apple ever said they are?), that they are not covering Adobe RGB gamut, so not good for still work (same who said they are?). People have some strange believe that because it's Apple made it's matching pro needs and specs. Sorry, but that's those people problem. If you read fine print Apple doesn't lie much. People simply assume too much. Next time don't trust Apple, just wait few weeks for reviews and measurements. It's really not that hard to avoid all frustration, just be patient.

This is random unit of latest 14 Pro:

Image
https://www.notebookcheck.net/The-new-M ... 327.0.html

as for home device I'm very satisfied with it. It's at least as good as high-end TVs out there when it comes to P3 coverage. Accuracy is crazy good as for random unit.

I see no lie in Apple latest claims for their HDR screen at all when it comes to brightness.
XDR is very capable peaking at 1600 nits and has 1000nits profile as well if I'm correct (+ proper calibration). Again as for home device it's very decent (blooming will always be there in zoned displays).

Latest Pro model are the same- nice 1600 peak. Not sure if Apple will have 1000nits limited profile, but going up is not that difficult (I assume way easier then trying to tone-map to lower nits). Those screens been already measured and promised values are real. Some already complain that standard mode is only 500 nits bright. They assumed by reading too much into what Apple said (again) that it will be 1000 nits.
Apple is going into very good direction for post needs with their devices. Introduction of different refresh rates is big plus. No one talks about it, which shows that most those 'experts' have no real understand what is important and what not. No one talks that now you have 10bit preview in many apps, which is also big deal.
There is still a lot to be done, but post/video needs in home device been not pushed forward for years, so there is big list there.

You simply look at Apple device from the wrong point. Those are not PRO screens. Those are very good home screens. Sometime needs overlay and you can be quite well covered for pro needs, but sometimes not at all. Apple PR is for masses. You should be better than this.
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 11:35 pm



EDR is a home technology to map HDR content to screen capabilities, instead of trying to stick to fixed (based on outdated specs) values. It's all to use in the best possible way given screen capabilities (if screen is 300 nits or 500nits max, etc.).
Nothing wrong with it for home usage at all! Problem is people don't get it and again compare to strict HDR specs used in pro industry. Same story again.
Offline

Jason Conrad

  • Posts: 797
  • Joined: Wed Aug 16, 2017 3:23 pm

Re: Resolve 17 to Youtube HDR

PostWed Oct 27, 2021 11:46 pm

MacBook PRO. Right there in the name. What about that says “consumer-grade display?” It’s false advertising, and the disingenuousness affects other products, like Resolve, which have bent to the lie, a-la rec.709-A. The “A” literally stands for Apple. Fact.


Sent from my iPhone using Tapatalk
-MacBook Pro (14,3) i7 2.9 GHz 16 GB, Intel 630, AMD 560 x1
-[DR 17.0 Beta9]
Offline

JonPais

  • Posts: 441
  • Joined: Fri Mar 21, 2014 2:17 am

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 12:11 am

thefilmaddict wrote:I just edited and colored a short video in HDR using resolve 17, studio 4k mini and an ASUS 1000 nit monitor. The video looks great while editing. When I export it, it's no longer in HDR and looks terrible (crushed blacks, too much chroma and blown out highlights). Youtube does not recognize it as HDR after I upload it (usually when it does, HDR comes up on the side of the screen). I think that I am using the wrong delivery/render settings. Does anyone know the right ones to use? There are videos out there with suggestions, but none that I've run into that mention the latest version of resolve (the new version has slightly different settings). Thanks for your help!
This helpful guide explains the settings required for uploading to YT.

https://daejeonchronicles.com/2021/03/0 ... 17-part-i/
https://daejeonchronicles.com
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 12:30 am

Jason Conrad wrote:MacBook PRO. Right there in the name. What about that says “consumer-grade display?” It’s false advertising, and the disingenuousness affects other products, like Resolve, which have bent to the lie, a-la rec.709-A. The “A” literally stands for Apple. Fact.


Sent from my iPhone using Tapatalk


We are talking about different pro :D Apple’s pro is just PR term.
Professional screens etc. used in studios.

Rec.709A is not directly Apple’s fault either.
Apple’s ecosystem has no problem to cover all current video needs. It’s only once you leave OSX ( YouTube, Vimeo etc. ) when your problems start. It’s all due to fact that industry failed to introduce up to date video flagging for HD format (BT.1886). There is basically no standard tag to say my video is Rec.709 with 2.4 ( or 2.2). It’s all that simple.
I written about it many times, so not going to repeat myself.
Offline

JonPais

  • Posts: 441
  • Joined: Fri Mar 21, 2014 2:17 am

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 12:51 am

Jason Conrad wrote:MacBook PRO. Right there in the name. What about that says “consumer-grade display?” It’s false advertising, and the disingenuousness affects other products, like Resolve, which have bent to the lie, a-la rec.709-A. The “A” literally stands for Apple. Fact.


Sent from my iPhone using Tapatalk
What is factual is what Andrew Kolakowski wrote: the displays on MacBook Pros leave little to be desired; and the machines are universally acclaimed as being among the very best on the market. Perhaps you confused WCG with HDR - but no one we're aware of ever has - and you are being dishonest with yourself when you accuse Apple of false advertising with regard to wide color gamut. If only Apple could take credit for what is inarguably one of the best NLEs in the world, that would be incredible - but I'm afraid the credit goes to Blackmagic Design alone. Doesn't the Internet have enough rancor already, without adding to it?
https://daejeonchronicles.com
Offline

JonPais

  • Posts: 441
  • Joined: Fri Mar 21, 2014 2:17 am

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 1:57 am

With Resolve Studio 17.4, color management has been greatly simplified, and I can't imagine an easier solution.

Screen Shot 2021-10-28 at 8.54.09 AM.png
Screen Shot 2021-10-28 at 8.54.09 AM.png (227.58 KiB) Viewed 3498 times
https://daejeonchronicles.com
Offline

thefilmaddict

  • Posts: 129
  • Joined: Sun Oct 27, 2019 9:58 pm
  • Real Name: Jason Morris

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 2:06 am

Thanks, everyone for your input. I've got to jump back on a SDR gig. I'll try these suggestions on Monday. I appreciate it!
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 9:09 am

Those settings ( simplified 17.4 HDR preset) and also one from link above are not correct.
If you use them you grade to full Rec.2020 gamut which will have saturation variation depending on viewing device ( how much it covers Rec.2020 gamut which can vary a lot).
Proper setting needs to limit gamut to P3.

These are correct settings:

Image
Offline

RikshaDriver

  • Posts: 641
  • Joined: Sun Aug 12, 2018 10:08 am
  • Location: Melbourne
  • Real Name: Asim Siddiqui

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 9:34 am

The biggest issue with RCMv2 Color Management is that Input DRT is still enabled by default.

This is an absolute no-no if one is looking for any accuracy from source footage as the signal gets altered whether you like it or not.

The HLG input bug with DaVinci DRT still hasn't been fixed either.
GitHub Projects: https://github.com/xtremestuff/
Commercial Plugins: https://xtremestuff.net/store/
Offline

Andrew Kolakowski

  • Posts: 9210
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Resolve 17 to Youtube HDR

PostThu Oct 28, 2021 9:38 am

This is the problem with presets (if you don't look and understand what is inside).

Return to DaVinci Resolve

Who is online

Users browsing this forum: dirk-pel, iconnel, Lucius Snow, panos_mts, Tettac and 203 guests